JP4345014B2 - Measurement and stabilization system for machine-controllable vehicles - Google Patents

Abstract

Device for measuring the movement and or position of an aircraft (1) comprises imaging optics (2) with which a sequential images of the environment (4) is obtained. The images are analyzed using an optoelectronic displacement sensor (3), which comprises a multiplicity of photosensitive surfaces and evaluation electronics for detecting the displacement of images structures and for outputting a measurement signal for the displacement. An independent claim is made for a method for measuring the movement and or position of an aircraft with at least six degrees of freedom.

Description

  The present invention relates to an apparatus and process for measuring the position of a motion-related measurement object such as speed and / or a machine-controllable vehicle for at least one of six degrees of freedom, where the term “vehicle” is optical Includes flying objects that use sensor technology. Furthermore, the invention relates particularly to its use as a stabilization system for controlling unmanned or manned flying objects. The purpose of such stabilization is to greatly simplify or replace control.

  Such stabilization is particularly important in remote controlled helicopters because it is difficult to control and can only be trained pilots. In the case of a helicopter, the control may in particular include one or both horizontal components of the flight movement. A further possible field of application of the present invention is the simplification of control of manned airplanes, especially near the ground and at landing. As further embodiments, there may be other applications such as distance measurement, ground altitude unevenness scanning, adjustment of flight motion between several flying objects, and the like.

  Conventional remote control helicopters only have a piezoelectric rotation sensor that stabilizes the movement around the vertical axis by the tail rotor as a stabilizer when used for hobby purposes or aerial photography purposes. Does not stabilize.

  DE 69502379 and JP-A-10-328427 describe adjustment control for a helicopter, in which a horizontal axis is artificially formed using a device for gravity acceleration and angular velocity existing therein, and a rotor blade Is controlled.

  In addition to the above, U.S. Pat. No. 5,738,300 measures and controls cruise speed relative to airflow (airflow velocity). In all the cases described above, control over the surface position and in particular stationary hovering is not possible.

  In the stabilization systems described in RU 9300250 (Russia) and DE 69507168, the direction of incident light is measured and evaluated by several light sensors in order to obtain information about the tilt and control it. However, this may also include detection of an approaching object. Both systems can only operate accurately when there is uniform illumination and a very uniform horizontal axis.

  EP 0780807 describes an autopilot that uses conventional equipment, such as a gyroscope, to stabilize flight attitude and horizontal speed in a relative manner. However, absolute measurement or adjustment control of the absolute position with respect to the ground surface is impossible.

  Properly maneuver the rotor blade control, for example, to control the helicopter's flight path, for example to perform static hovering, first the tilt (ie roll and pitch angle) and secondly the resulting speed. Need to be controlled by. For this, it is necessary to know the tilt as well as the horizontal velocity (preferably with respect to the ground surface). In the case of autopilot flight, it is not enough to respond to the resulting tilt. This is because even in this case, the helicopter probably does not slow down its given speed. Rather, it must be stopped by a properly adjusted tilt in the opposite direction.

  From this point of view, the object of the present invention is to measure and stabilize the horizontal movement with respect to the ground surface.

  DE 69426738 T2 describes a navigation system having an image sensor mounted on a gimbal mount, the image data is analyzed for shift motion and integrated to determine the position.

  DE69310695T2 describes the evaluation of images obtained at a flight position by one or more cameras. A specific characteristic image part was continuously analyzed with respect to its optical movement. From here, the center or vanishing point of these movements is determined, and the detected shift speed is analyzed and weighted by the distance from the vanishing point, thereby calculating the distance to each ground point for navigation purposes.

  DE 69012278 T2 also recognizes this for navigation purposes, comparing it with a map of the currently taken image and a previously stored image.

  All three methods described above require complex image processing and at least one camera. Measurement speed is subject to camera specific frame rate constraints. In video cameras this is inconvenient for fast attitude control, especially in the case of small helicopters. For a remotely controlled lightweight flying object, such a method is disadvantageous due to its weight, cost, and scanning speed constraints.

  Position measurement by satellite navigation (GPS) has the disadvantage that it cannot be detected with sufficient accuracy even though micromotion is important near the ground. Furthermore, due to weight and cost reasons, application to a simplified model is not reasonable.

  The object of the present invention is a system of the type described above, which operates without the above-mentioned disadvantages and accurately detects movement, and can therefore be applied as a measuring device for stabilizers.

  As a solution, the features described in the independent claims are provided.

  A further object of the present invention is to stabilize the control of the flying object. For this purpose, the features described in the dependent claims 18 and thereafter are provided.

  The optical image processing system may be a group of, for example, a convex lens, a camera lens, a concave mirror, and any other image projection lens.

  As the shift sensor, an optical sensor having many photosensitive elements having low sensitivity compared to the image acquisition device and integrated with an electronic evaluation circuit on the same chip is used, and the sensor is a motion sensor. And is commonly used for optical mice. Hereinafter, the latter is referred to as an optical mouse sensor.

  Such a shift sensor may include a detection area on a substrate, which is composed of a plurality of photosensitive areas (pixels) in which signals are often read out, and the reading speed may be significantly higher than the frame speed of the video camera. Furthermore, an evaluation unit is included on the same substrate, in which the pixel signal is analyzed electronically with respect to the transition. Typically, the transition is evaluated sequentially with respect to two orthogonal coordinates and can be output as separate values (delta x, delta y). The shift can be broken down into smaller steps, which may be equivalent to pixel distance, for example. Sensor detection may include both the direction and magnitude of the optical shift (translation).

  Commonly used optical mouse sensors can recognize transitions even when the image structure of the mouse pad is randomly and irregularly distributed, such as for example, microscopic sized fibers or paper pieces. , Usually 16 × 16, 18 × 18, or similar number of pixels in CCD technology (charge coupled devices). The transition values can be evaluated by comparing the data of these pixels with respect to their own coordinates by reading them continuously and frequently to obtain a correlation with the previous time of the same sequence. . This comparison may be performed using the most recently read data or using data from earlier cycles of the same sequence. Data can be temporarily correlated with each other. The evaluation process can be performed digitally under program control. A similarity algorithm may be used. For this purpose, the luminance signal of an individual pixel is first divided into several stages of differentiated luminance, thereby reducing the amount of information to be analyzed. Includes adaptation to average image brightness. That is, the pixels are analyzed with respect to at least one shared reference value that varies in shutter time and / or matches the average brightness. This type of method allows the evaluation process to function correctly even when the background or the brightness of the outside world changes.

  The shift sensor can be suitably used as a sensor according to the present invention even though it has been originally used only for application to an optical mouse and only for navigation on a computer screen. That is, it is combined with the above-described optical means for imaging a remote object. As a result, application to scanning of a remote object in free space is possible.

  In many cases, illumination of objects is obtained exclusively using ambient light.

  Further advantages of a sensor having a photosensitive array and its electronic evaluation circuit on the same chip are low cost, light weight, and high evaluation.

  The advantage of a relatively small number of pixels is that the scanning speed is faster compared to systems that require television cameras and video cameras. The smaller the flying object, and thus the more quickly it responds, the more important the sampling rate is in order for the flying object attitude control to function sufficiently quickly. By eliminating the large amount of data in the video signal, the advantage of improved processing speed can be obtained. The shift sensor can operate at a clock frequency lower than the clock frequency specified according to the data sheet. Thereby, the exposure time can be lengthened and thus higher photosensitivity can be obtained. Decreasing the clock frequency to (original) 75% to 10% or less can result in a data rate substantially faster than with a video camera.

  In contrast to the application of a shift sensor in an optical mouse, the device according to the invention must be designed to target fairly large and distant objects. However, in addition to the optical equipment used in the optical mouse, in cooperation with the current lens, it is possible to provide a complementary optical equipment that is dimensioned to perform the imaging defined in the independent claim. it can. This can be done, for example, by adding a concave lens according to Barlow's principle or by projecting an image onto an intermediate image plane.

  As an advantage, the imaging optics system may focus at infinity, or more preferably at a distance that results in a sharpness range from infinity to the shortest possible object distance, including depth of sharpness. Can be combined.

  The focal length can be selected according to the required angular resolution based on the spatial resolution defined by the sensor. Focal length selection can also take into account the maximum depth of field that can be obtained (range of depth of field) and the maximum measurement range for speed. The focal length suitable for application to a remotely controlled small helicopter is in the range of 7-25 mm, preferably 10-12 mm. This provides a sufficient angular resolution, a sufficient velocity measurement range, and an appropriate depth of sharpness ranging from about 20 cm to infinity. Here, the term “focal length” means not a distance to the image plane but a lens characteristic.

  The sensor unit (5) obtained by combining the imaging optical system (2) and the shift sensor (3) according to the present invention has a conical search range reaching free space outside the device, and its angular width is It is defined by the size of the total photosensitive area of the shift sensor and the focal length of the imaging optical system. Unlike systems that use imaging equipment such as video cameras, most of the following description ignores the angular width of the search range, and the imaging optical axis or main direction (hereinafter referred to as “line of sight”) (11) Only consider.

  In accordance with the present invention, the structure and contrast of the atmosphere, the ground, and other distant objects are optically imaged onto the shift sensor. In general, the texture required by the above-mentioned shift sensor to detect a transition arises from any contrast characterized by the outside world or part of the terrain due to surface texture, details, contours and similar features. Also good. These may arise, for example, separately from the boundary of the object, or the optical roughness of its surface, similar to the operation of an optical mouse. Almost all visible structures that normally occur in ground images contain such contrasts, structures, and textures so that the shift sensor described above can detect image translation even in flight. Exceptions are water, fog or continuous snow.

  Although the following items a) to d) describe that the movement can be measured by a device mounted on a vehicle, for example, a flying object, the present invention is not limited to this.

  In general, both vehicle rotation (a) and translation (b) can be measured. This is because both types of movement cause movement of the virtual image into the imaging optical system in the direction of light incidence.

  Due to imaging, the angle of incidence is converted into a proportional image shift on the shift sensor and can be measured in this way. Thereby, the focal length of the imaging optical system is a proportionality factor.

  Depending on the configuration and design of the present invention, measurement objects with different vehicle positions and directions can be measured.

  a) When measuring the rotation of a vehicle, eg roll, pitch and / or yaw, a sensor device is mounted so that the line of sight or at least its vector component is transverse to the axis of rotation; Image shift is measured along the component orthogonal to the axis. In order to stabilize the flight movement and avoid unnecessary shaking, the control value is generated using the measured value by comparing the measured value with the target value, and the control signal is sent to the relevant control device to perform the correction. By doing so, closed loop control can be established. In order to stabilize the nose, suitable configurations are diagonally forward downward and / or backward downward.

  Furthermore, tracking can be realized by scanning the direction of the target. For this purpose, the closed loop described above is provided. Control values can be tracked. That is, control the aim or direct the course to the goal as a result of aiming. In both cases, the aiming direction of the scanning device is more influenced and under control of the self-regulating loop. This can be used for target tracking even when the object itself is moving. For this purpose, the line of sight is initially aimed at the target, and this aim is maintained by the adjustment control.

  Another application is tracking optical equipment with its own servo motor. This can be used to stabilize the camera for moving pictures and photos. Advantageously, several types of unwanted movement can be stabilized. That is, the first is the swinging motion of each vehicle or flying object, the second is the protrusion of the image section caused by the drive, and the third is the possibility of the target object itself moving. Conventional gyroscope stabilizers only compensate for the first type of effect. Another advantage is that, as described above, the response of the shift sensor is significantly faster than the frame rate of moving image recording. The device described above is also suitable for general use without being limited to vehicles. From this point of view, a sensor unit equivalent to the apparatus described in claim 1 is mounted together with a camera on a camera platform movable by a servo motor, and the sensor unit moves together with the camera.

  b) In order to measure the translational movement, the optical line of sight is directed transversely to the direction of movement being measured, at least with the vector component. The amount of image shift detected is proportional to the distance traveled by the vehicle. The same can be applied to each speed. Furthermore, the image shift is inversely proportional to the distance along the line of sight, proportional to the focal length, and proportional to the sine of the angle between the direction of motion and the line of sight, depending on the shape of the ray.

  If only translation is measured, the influence of the rotation described in (a) may get in the way. The optical measurement signal may be purged from these effects. That is, a rotation signal representing rotation and a ft sensor signal are synthesized by calculation. The rotation signal can be easily obtained using another measuring device such as a gyroscope. Depending on the polarity, the computational composition may consist of subtraction, addition, or generally mixing. The signal may be processed in advance, for example by differentiation or integration. This compensation method eliminates the gimbal mount for the optical receiver used by the cited prior art (the latter is usually a video camera).

  When measuring the horizontal flight movement of a flying object, the line of sight may be fixed below with respect to the fuselage. Next, on the assumption that the robot is in a normal posture, the direction in which the light incidence of the detected ground texture can be measured deviates due to the horizontal flight motion. Therefore, for example, the horizontal position such as the ground speed or the ground position can be stabilized by the above-described closed loop control applied to the horizontal motion value. In the case of a helicopter, this can be used to stabilize hovering flight horizontally. That is, periodic rotor blade control is activated.

  As described above, the horizontal direction measurement value is inversely proportional to the viewing distance. To obtain an independent measurement, use a distance measuring device to measure the optical velocity measurement signal (optionally after being purged from attitude effects) at least within the sub-area of the possible distance range. You may multiply by the coefficient which increases continuously with increase. When measuring the distance, a microwave radar or ultrasonic radar known as a vehicle reverse collision sensor or an autofocus camera, or an optical distance sensor according to the optical scanner principle, for example as described in DE 4004530, is used. It's okay. However, this altitude compensation may be omitted if the measured value is used only for stabilizing loop control applications, or may be performed only in a limited range. Hovering stabilization in the helicopter was obtained using a PID control loop without altitude compensation. This functioned very accurately within a quotient ratio of 1:10 at any surface distance, depending on the loop amplitude adjustment, and even better when the ratio was within 1:30.

  If this compensation is omitted, the higher the measurement sensitivity, and thus the lower the flight altitude, the more reliably the control functions. If a fixed target value is given for the (uncompensated) speed, the actual speed is automatically decelerated near the ground. This behavior has the side effect that at that time, when the speed is reduced due to the conversion of kinetic energy, upward movement occurs simultaneously (except when the tailwind is faster than the ground speed). Thus, collisions are avoided by momentary ascent while cruising horizontally over the surface ridges. This behavior is also obtained by delaying the altitude signal used for compensation when altitude compensation is present, i.e. with respect to time.

  c) If the speed is approximately known, for example the distance to the ground surface or to an object can be measured as well. For this purpose, the optically measured signal is employed as the reciprocal of the distance. For example, a known speed can be divided by an optically measured speed value. When this value is used in the control loop, division can be omitted if the target value is already defined as an inverse.

  One application is the control of ground distance in airfoil aircraft. Since the horizontal velocity is approximately known and almost constant, the current flight altitude can be measured from the surface even when flying over a ridge. In this way, it is possible to detect an approach early and avoid a collision with a visible object. The flight altitude can be controlled or limited to the shortest distance from the ground. For this reason, it is possible to adjust the line of sight obliquely forward and downward to obtain a time margin. The distance can also be measured not with respect to the ground surface, but with respect to an object that is present there, for example forming a ridge.

  d) In general, motion can be measured and correlated in relation to various different coordinates. In such a method, it is preferable that the mixed measurement objects are separated from each other as shown in the following embodiments and FIGS. It is also possible to measure distances and distance changes along the combined line of sight. This makes it possible to measure each of the ascending speed or descending speed, or to recognize the approach.

  In this specification, various measurement objects are generally referred to as “movement-related measurement objects”. This can be applied to local values as well as angle values. The motion-related measurement object may be of a static type, such as position, distance, or posture, and / or a dynamic type of, for example, speed or acceleration. Therefore, the terms “motion”, “rotation”, “translation”, “tilt” etc. are always used here in general terms, ie with respect to position and their respective rates of change.

  Motion related measurement objects can be measured against the ground surface or against movement or other flying objects associated with the ground surface. According to the present invention, the output signal of the optical sensor, hereinafter referred to as “sensor signal”, can be utilized and evaluated in several ways, as described below. Optical mouse sensors often have quadrature outputs that send incremental signals on two lines for each of the two coordinates, with each shift being signaled as a step or multiple steps along its direction. Similarly, a serial data link or any other type of signal transfer may be used for data output. In many cases, each data read provides the number of steps of change that has occurred since the most recent read.

  By using appropriate evaluation values, various information can be obtained, for example, a) direction and speed or angular velocity of motion, that is, rate of change, b) range of motion or position or angle value, and c) acceleration. It is not limited.

  a) The rate of change is obtained by differentiating with respect to time. This is consistent with the frequency measurement of the incremental shift step taking place. For this purpose, in a sensor outputting a shift in the form of discrete incremental steps, the frequency of these steps, for example the frequency of the output quadrature signal, is measured.

  It is preferred if the speed measurements are obtained continuously or at least pseudo-continuously. Each signed sum for each interval of occurring steps is used in order to obtain an appropriate velocity measurement within a continuous time interval, preferably chosen small. For example, step addition / step subtraction or a total changed from the latest output is read, and each obtained total is divided by the length of the interval. Division can be omitted if the lengths of the intervals are always equal. Alternatively, in each step in which the signal changes, the time width since the changed preceding step can be obtained, the reciprocal thereof can be obtained, and a sign can be given corresponding to the direction of change.

  Furthermore, the frequency values obtained in this way can already be recalculated before the next change occurs. For this reason, after the period corresponding to the frequency has expired, if the frequency value is reduced continuously or stepwise, the reciprocal of the currently expired waiting time is less than the most recent frequency value. The value to be

  b) The position value is calculated by adding / subtracting or summing the incremental steps of the sensor signal, eg counting the steps corresponding to the direction or integrating the increments, similar to the normal operation of a mouse on a computer screen. Obtained.

  Alternatively, or in combination with this, the position can be obtained by continuous integration with respect to time of the frequency value obtained as described above. This makes it possible to compensate for the influence of other movements on the measurement values mentioned above prior to the integration process, so that the integration results are obtained in a form already compensated.

  Depending on the embodiment, various measurements related to the position measurement are consequently possible. Using a line of sight directed downwards, position measurements associated with the ground surface are obtained.

  The measured values from the device according to the invention can be used to establish a control loop, and the control values are generated using the measured values in the adjustment controller (7) in the same loop to control the movement. For this purpose, a method known as PID may be used. The actual measurement may be any measurement generated according to the present invention, i.e. velocity related, position related, or partially mixed. In addition, single or multiple integral and derivative values may be generated and included. For example, acceleration values can be obtained by iteratively differentiating frequency measurements. Further, PID control may include the use of proportional, derivative and / or integrated portions of the signal from other onboard instruments such as, for example, piezoelectric gyroscopes.

  Manual control can be supplemented and stabilized by superimposing the manually supplied control signal on the output signal (control signal) of the adjustment control circuit. Manual control signals may also be included in the regulation control loop as target values, for example by mixing them with the inputs. Thereby, a comparison between the target value and the actual value is obtained, in which case the target value is derived from a manual signal. A detailed description of the adjustment control circuit is given below. See the first embodiment and FIG.

  For remotely controlled flying objects, the parts required for control may fly with the object or may be located on the ground and connected by wireless communication.

  As the shift sensor, other optoelectronic devices can be similarly used instead of the optical mouse sensor. The shift sensor includes at least two adjacently arranged photoelectric light receivers whose mutual distance is a quarter of the spatial frequency wavelength that is numerically present in the image of the structure on the ground. Corresponds to or corresponds to a size corresponding to one. The motion of the optical structure induces an alternating signal that is phase shifted in the optical receiver. These alternating signals are fed / induced to a circuit that analyzes the time difference and / or phase difference. This type of circuit may be a time comparison circuit because it is used with a direction sensitive photoelectric relay or phase comparison circuit. This type of circuit detects, based on the phase position, which of the two signals is ahead of or behind one, thereby detecting the direction of motion and speed with a certain degree of accuracy. Furthermore, the circuit may support incremental analysis according to the orthogonal principle and thus may be configured in analog or digital form. On the other hand, since the captured part of the ground structure is smaller, the measurement accuracy of such devices is usually less than that of a shift sensor with an array and integral analysis electronics.

  The optical image can optionally be amplified by an afterglow amplifier or other insertion optics before reaching the shift sensor.

  All of the signal processing and computing operations described above can be implemented, for example, digitally in a program-controlled microprocessor or in analog circuitry. Some of the partial methods described herein or set forth in the claims may be combined with a single combinatorial process and / or performed by a single shared processor.

  The above procedure can also operate with infrared light. Thus, the terms “viewing”, “light” and “imaging / imaging” include all kinds of light radiation.

  Each device may be combined with a light source to be applied in the dark. Preferably, it is emitted in a guided manner and directed towards the sampling object.

  Hereinafter, the operation will be described in detail with reference to an example of the embodiment.

  FIG. 1 shows an example of a first embodiment in which the horizontal movement of the helicopter (1) is measured and stabilized. The lens (2) functions as an objective lens and projects the visible section of the ground surface (4) onto the optical shift sensor (3) equipped with an integral analysis device. During the forward movement, the ground point (44) optically moves to position 44 ', thus shifting the image on the sensor (3). The sensor outputs an incremental shift measurement in two coordinates (x and y). As shown in the figure, when the line of sight is downward, information on both the traveling direction and the lateral direction regarding the flight speed with respect to the ground surface is obtained.

  These measurements can be used to stabilize the flight path or stationary hovering flight. For this purpose, the feedback control electronic component (7) is connected to the optical measurement signal on the input side and outputs a control value for controlling the flying object on the output side. The electronic component includes at least a part of the overall control function, in which case the periodic adjustment of the rotor blade (8) controls the inclination of the main rotor plane, for example by means of a rocking plate, and accordingly Affects the horizontal acceleration of the helicopter. An example of the operation of the adjustment control (7) will be described in detail below with reference to FIG.

  On the other hand, the measured value depends not only on the horizontal motion but also on the rotational motion simultaneously performed by the flying object such as rolling and pitching. This effect is compensated by mixing the rotation signals. The rotation signal may be proportional to other rotation values such as angular velocity of rotation or angular position, for example. Different methods for obtaining a rotation signal adapted for compensation can be described below.

  The rotation signal can be obtained by a gyroscope or a piezoelectric rotation sensor (9), (10) (piezoelectric gyro). Such an additional measurement pickup can be provided on the flying object or can be attached or mounted on the device (5) according to the invention. Since piezoelectric gyros measure angular velocity, compensation is reasonably done by mixing the gyro signal with a frequency signal already representing velocity, for example obtained from a shift sensor by differentiation.

  In the case of a helicopter, a rotation signal may be generated. That is, the acceleration sensor moves with the main rotor shaft, is placed at a fixed distance from the rotor blade shaft, the acceleration component parallel to the rotor blade shaft is measured, and the continuous measurement value is a function of the rotation of the rotor blade. As described above, the phase is periodically analyzed. As with the gyroscope, rotation of the rotor blade plane induces a forward force that occurs and is measured as a periodically varying acceleration in the acceleration sensor. The amplitude of the change is a measurement value of the angular velocity of the rotation to be measured. The phase position relating to the rotation of the rotor blades is a measurement value of the direction of rotation and the direction of the axis to be measured. Phase-related analysis is performed by generating a sampling sequence that is synchronized with respect to the rotor blade rotation, such that the measurement signals cross, or are switched from quadrant to quadrant, for example, according to the measured clock rate. Thereby, the measured value of rotation is obtained by being divided into components such as a pitching axis and a rolling axis. The synchronization can be performed by a rotary encoder that samples the rotation of the rotor blade shaft.

  Preferably, the measured force occurs as a periodic change. Therefore, the acceleration sensor need not be able to measure absolute values, but only need to be able to measure changes. Therefore, a piezoelectric converter with a price suitable for the budget can be used as the acceleration sensor, and the measured value can only be capacitively coupled. Furthermore, as a result, there is an advantage that no zero error of the measured value occurs after the start, ie the measured value is obtained in the form of amplitude. Signal transmission from the rotation sensor to the regulation controller can occur by wireless communication, optoelectronic transducer, inductive signal coupling or sliding contact. Even in the absence of optical measurements, this measurement procedure can generally be applied to the stabilization of a rotorcraft whose slope is adjusted and controlled according to the measured values. For this purpose, the acceleration sensor moves with the main rotor shaft, is placed at a certain distance from the rotor blade shaft, the acceleration component parallel to the rotor blade shaft is measured, and continuous measurement according to the position of the rotor blade It is sufficient that the values are periodically analyzed and synchronized with respect to the phase.

  To obtain a rotation signal to compensate for the effects of rotation without using a separate sensor, a control signal can be used as the rotation signal, and a control signal, for example a servo signal for the tilt of the rocking plate, controls each rotation. Is supplied to the control means. This works because the flying object usually responds with a roll or pitch speed that is very accurately proportional to operation. Such a control signal thus provides a suitable measurement of the angular velocity used to compensate the rotation signal. In so doing, compensation is performed by mixing the control signal with the velocity signal obtained from the shift sensor, which can be obtained by frequency measurement. When the control signal itself is generated according to the present invention by an adjustment control loop, the actual value includes the measurement signal to be compensated, and the compensation is the inverse of the negative feedback type adjustment control loop from the control value to the actual value. Adapt to mixing to reduce loop amplification. In this case, according to the same principle of operation, compensation mixing of the rotation signal can be realized simply by means of a regulation control loop with a small amplification each.

  As a possible alternative or combinatorial means of obtaining a rotation signal, it is to use a second optical shift sensor for this purpose, which functions as a rotation sensor according to the invention as described above, whose line of sight is the first It is aimed differently from the sensors in

  FIG. 5 shows a block diagram of the adjustment control device (7) for the helicopter (group). The adjustment control loop may be configured similarly to the pitching and rolling axes and is therefore only shown once. During normal manual operation, control signals are output to the rotor blade control 24 via the receiver 21. The position reached in flight is the result of the flight response behavior and matches a series of several integrals over time, as shown in a self-evident manner. All these motion-related values represent the actual flight motion at each time point. The effect of movement on the measurement is indicated by a broken line. Under the measuring instruments (9/10), (25), (3), the operation of the adjustment control device (7) is shown separately for the measurement object and the expression of the adjustment control itself. And the target value is compared.

  The frequency (27) is first obtained from the signal of the shift sensor (3). The frequency can be purged from tilt and possibly flight altitude effects as described above. Here, this is performed by addition (24). Subsequent integration (28) may be summing or counting and produces a position signal, with the derivative effect of the frequency measurement reversed again.

  A target value for the flight path or flight speed is given. This may be, for example, a velocity vector or a position vector parameterized as a function of time. This function may be programmed prior to flight or pre-set during flight, or may be zero (as a velocity vector) or constant (as a position vector) in special cases such as hovering. Good. The current target value is generated, for example, by the wireless communication receiver 22 and compared with the optical measurement value by subtraction (30a). The obtained difference corresponds to each instantaneous deviation from the target route (30a) or target speed. In the case of the required position vector target value (illustrated), the adjustment occurs because the position deviation (30a) is first defined as the control value for the flightback speed, and this speed itself is adjusted and controlled. (30b). According to the PID method, the mixed value of the position measurement value and the speed measurement value may be adjusted and controlled to each other.

  An inclination control value (30b) is defined in proportion and inversely proportional to the speed deviation, and the inclination of the main rotor blade plane is controlled accordingly. When the helicopter tilts, it responds with horizontal acceleration proportional to the tilt. Accordingly, the flight speed changes in proportion to the time integral of the slope, and the adjustment control loop is closed.

  In order to bring the inclination closer to the target value, the actuator of the swash plate (8) is controlled. Since the helicopter cannot directly control its inclination, but only through the speed of change (angular velocity), the inclination is not directly known from the position of the actuator. Therefore, it would be advantageous if its own measurements could be used to control the tilt. In FIG. 5, a tilt sensor (25) is prepared for this purpose, and its own target value comparison (30c) is possible. The result is supplied as a control value (26) to the rotor control function (8) via a mixer or switcher (23).

  As the tilt sensor, for example, a device according to the cited RU9300250 may be used. But this is not accurate enough. Alternatively or additionally, the tilt of the rotor blades can be represented according to the operation of the artificial horizontal axis by introducing a time integral of the signal of the piezoelectric gyroscope (9, 10) indicating the angular velocity of the tilt. However, the problem is that the integral calculation produces an undefined integral constant corresponding to the unknown slope of the horizontal axis. This can result from slight drifts as well as switching-on conditions and greatly hinders control.

  The measurement of the tilt is made by differentiating the measurement signal (24), which is obtained from the optical shift sensor (3) and is proportional to velocity, with respect to time in at least part of its frequency range. This works because helicopters in normal flight conditions accelerate in proportion to their inclination, and the acceleration is obtained by differentiating the measured speed.

  As a result of the differential operation applied to the optical measurement signal output in a stepwise manner, a discontinuity that becomes a disturbing factor may occur from the incremental step. The measured value of the tilt can be generated by combining both the measured value obtained from the shift sensor and differentiated with respect to time, and the time integration signal of the rotation signal proportional to the angular velocity of the tilt. In this respect, in particular, the weighting of the spectral components of the high frequency differentiated measurement values may be small and the weighting of the spectral components and constant signal components of the low frequency integrated signal may be small. In this way, if the low spectral component is extracted together with the discontinuity, it is smoothed by lowering the high spectral component, so that the above-mentioned problems concerning the integration constant and the zero point drift do not exist and are solved. Each missing part can be replaced by another signal. Combinations that occur in the optical measurement signal as a result of differentiation and low-pass filtering can also be expressed and generated as a first high pass in a synonymous manner, and combinations that occur in the gyro signal as a result of integration and high-pass filtering Can also be expressed and generated as a first-pass low-pass in a synonymous manner.

  As a rotation signal for tilt adjustment control, for example, each method described in relation to rotation compensation (according to claim 23) such as a piezoelectric gyroscope or a control value for control may be used. Alternatively or in combination, the rotation signal may again be generated as described above.

  Tilt adjustment control can be combined with horizontal motion adjustment control. That is, the tilt and motion measurements (velocity and / or position) described above are mixed and thus adjusted and controlled together. In order to deepen the understanding, the operations of “representation of measurement object” and “regulation control” are clearly distinguished, but it is not essential for implementation. In general, the proportional, differentiated, and possibly integrated parts of the optical measurement signal are mixed in parallel to the regulation control loop, so that the overall regulation control described above is synonymous in a shared PID regulation control loop. It is realized in the form of

  In the use of the integral signal branch in the regulation control loop, the stationary regulation control is obtained, so that the hovering flight is stabilized and restored to its original position even after a temporary shift such as a disturbance, gust.

  In the integral calculation of the shift value, a measurement value proportional to the position can be obtained. When obtaining an absolute spatial position in ground coordinates, yawing motion can be considered in a compensatory manner. For this purpose, the incremental sensor signals of both sensor coordinates can be summed by integral calculation, providing two integral sums representing the two ground coordinates, and with respect to the rotation axis orthogonal to the ground coordinates. The sensor sequence is measured and the sensor signal increments to be summed are vectorized under control of the measured direction of travel before being summed.

  A target value for speed, not position, can be provided. Such a target value is first integrated and can then be used as a position target value. The actually measured value and the target value proportional to the position can be omitted, and the adjustment control can be limited to the speed. Mixing (30a) and integration (28) can be omitted and a manual control value (22) can be fed directly to the speed comparison function (30b).

  The adjustment control method described above, particularly the method of adjusting and controlling the inclination of the helicopter, can also be applied to flight stabilization without using an optical sensor. Thereby, the actual measurement value and the target value proportional to the position can be omitted, and the adjustment control can be limited to the tilt angle.

  With an alternative or combination of optical sensors (3), any navigation system known in the prior art can be applied to obtain the actual position or velocity (28), thereby adjusting as described herein. A control method can be realized. For this purpose, the adjustment control method described above omits the sensor (3) and the frequency measurement function (27) and adopts the position measurement (28) and / or velocity measurement obtained from an alternative measurement system. Just change to.

  FIG. 2 shows an example of a second embodiment according to claims 36, 37, which includes at least two shift sensors (31), (32) for measuring rotational movement or rotational speed. Both sensors may be located at a distance behind the shared lens (2) or other imaging optical system, or may have their own imaging optical system, as shown in FIG. . Thereby, the resulting configuration may be an integrated device or may be assembled in a separate unit. The lines of sight of different sensors are aimed so as to branch at a predetermined angle. For most applications, an acute angle is appropriate. Thus, both lines of sight are similar and the main line of sight is eventually shared, which can be considered as an angle bisector. As a result of rotation about the bisector of the angle or about an axis near it, an image shift of the tangential ground point image (61, 62) occurs. Therefore, the sensors (31, 32) are aimed in a manner that detects a tangential shift. The tangential shifts measured by both sensors may be related to each other, for example by subtraction, mixing, comparison or generally superposition of the measurements of both sensors. In this way, the tangential portion of the shift is removed by filtering, and the measured value becomes substantially from other movements. The signal comparison may generally consist of mixing or superposition. If the individual measurements are weighted differently before mixing, the position of the axis of rotation can be changed in a defined way. At the same time, different measurements may be obtained from other mixtures of measurements, for example according to the example of the first embodiment, the sum of equally induced movements can be used. Instead of superposition, any known type of signal comparison may generally be used.

  FIG. 3 shows an example of a third embodiment according to claims 36 and 38. It is adapted for use to measure distance or distance change or approach speed. The configuration shown in FIG. 2 is used, but instead of a tangential image shift, the radiation component of the image shift is measured. Subtraction of both sensor signals, or generally superposition of both measurements, filters out the radially backpropagating portion of the shift. As each object or ground surface (2) approaches toward position (4 '), the visible point (41) moves to position (41') and its image on the sensor (31) moves to position (51). ) To the radial direction (51 ′). The same is true for sensor 32 for point (42, 42 ') and image (52, 52'). Due to the beam shape, the measured angular difference moves in proportion to the change in distance and inversely proportional to the square of the absolute distance. If the distance is approximately known, the change can thus be measured and the speed in the direction of travel can be measured starting from the change signal.

  In contrast to the distance measurement according to claim 31, this is not the absolute value of the distance to be measured here, but the rate of change. Furthermore, here different speeds exist in the lateral direction and need not be known. If the line of sight is in the vertical direction, the resulting rate of descent or ascent will be measured.

  On the other hand, when the speed is approximately known rather than the distance, the absolute distance can be determined by the same configuration. Since the measured value is the square of the reciprocal of the distance, the accuracy of the measured distance is twice as good as that of the known speed method. For example, if the speed is only known up to ± 10%, the measuring distance will therefore change only 5%.

  All the above measurement methods for distance or approach speed can be applied to collision detection, for example to anticipate or avoid collisions.

  Apart from the sum of the radial shifts, a difference may be formed, i.e. the sum of translations in the same direction. Thus, the lateral movement about the other two axes can be measured simultaneously.

  The rotation and approach measurement methods described with respect to FIGS. 2 and 3 can be combined. That is, the measured values of both coordinates of each of the at least two shift sensors are analyzed.

  FIG. 4 shows an example of the fourth embodiment. Three sensors (31, 32, 33) may be arranged in three different directions, for example perpendicular directions, each sensor having two measuring directions. As a result, a maximum of six measurement values are obtained. The line of sight may be a right angle, but it is not essential. In order to stabilize the flight, all three lines of sight may be directed to the surface like a tripod. The measured values of the oblique line of sight can be easily converted into linear coordinates related to control by calculation mixing according to the rotation matrix. With such a configuration, the orientation with respect to all six spatial coordinates can be obtained, and all components of the flight motion can be detected and stabilized. All measured values can be purged due to the above mentioned effects caused by rotation. That is, three independent rotation sensors are provided. The rotation axes of these sensors may be aimed at the same vector system for simplicity.

  However, these independent rotation sensors may be omitted. To compensate for one of the sensors (31), (32), (33), one or both signals of the remaining sensors are used, so that these remaining sensors are used for rotational measurement, The measured value is used as a rotation signal for compensating the first sensor according to claim 23.

  In general, any of the measurement methods described herein can be combined. In addition, the effects of movement related components interfering with the measurement of other movement related components are thereby eliminated. That is, another mixture of movements is measured with other sensors and both signals are compared. Motion-related values can be measured with some degree of freedom independent of each other, even when these degrees of freedom comprise a combined motion, ie, motions that are not separated from each other. Each motion or coordinate axis need not be orthogonal to each other. Each sensor can separate measurements into separate coordinates to measure other mixtures of motion related components. That is, the measurement signal is converted to a linearly independent and optionally orthogonal measurement signal by appropriate mixing according to the laws of vector geometry.

  In general, any number of sensors may be applied and these may be aimed differently. In addition, sensors that provide sufficient contrast can be selected to analyze the signal. This reduces the probability that the image details are insufficient.

  In performing the adjustment control, the above-described conversion is not essential. In general, the spatial coordinates of motion can be measured and adjusted in a mixed manner as well. With appropriate sensor aiming, the intended mixing can be defined. In the case of helicopters, optical measurements of rotation can be applied to control the tail rotor. Thereby, the line of sight may be directed backward and obliquely downward. The combination of vertical axis rotation and lateral drift is then measured. In this way, the aircraft aligns itself in the direction of flight, independent of the wind.

  Other aeronautical equipment already commonly used for stabilization can be substituted as well. For example, in the case of a remote control model, a commonly used vertical axis gyro can be replaced with a rotation measurement. Furthermore, instead of a variometer, an optical measurement of the ascent rate as described in claim 31, 38, 39 or 40 may be used.

  In combination with instruments commonly used to measure other motion-related components, such as flight altitude (barometer) and nose alignment (compass), complete autopilot can be achieved, completely replacing control.

  For GPS controlled applications, accuracy, resolution, and measurement speed can be significantly improved by supplementing with optical measurements, especially in the range near the ground.

  If the illumination or contrast coming from the surrounding environment is insufficient, the sensor signal may be incorrect. Most of such errors are undersized or completely missing, and most appear in the form of shift values that occur locally, that is, at a single location. In order to increase the number of usable image positions and to reduce dropout, as already described with reference to FIG. 2, 3 or 4, several sensors with different line of sight are provided to provide sensor signals with respect to each other. Can be analyzed. As a result, a shared measurement result is obtained, and a sensor having a smaller measurement value is not weighted or weighted. The term “smaller” refers to the magnitude of the shift measurement from one or more coordinates and / or the actual brightness output from the shift sensor and / or the contrast quality measurement of each sensor. It's okay. The shared measurement value can be obtained by, for example, maximum value formulation, weighted average formulation, or switching to each strongest sensor.

  An array can be generated from a plurality of shift sensors, and the sensors can be interconnected in a matrix network to obtain a weighted result as described herein. Further, it can be combined with signal analysis methods as described with respect to FIGS. This allows a thorough analysis of motion and is far superior to methods based on analysis of video signals as far as speed is concerned.

  In order to avoid control errors that could interfere with the measurement when the measurement signal is omitted, the control can be switched to another mode independently or in combination with the methods listed above. In this case, no optical measurement signal is required. Switching can be controlled by the absence of the measurement signal and / or below the minimum contrast value and / or minimum brightness. Switching of the adjustment control is executed. There, the weighting part in the mixing of the actual value and the measured value to be mixed, and a method in which the adjustment control is appropriate, for example, with a configuration known from the prior art, so that the optical signal exists again sufficiently without a shift sensor It will be reorganized until. In its simple and special case, the tilt of the rotor blade can be kept horizontal while the measurement signal is missing.

  A device as described above or as claimed may also detect or measure relative movement between the device and any kind of object located at a distance therefrom (( Can also be used for measuring the angle change in the direction of the optical tilt of the virtual image)). Thereby, the optically sampled object may be moving on the one hand and / or on the other hand, the object may be equipped with a sensor device, similar to the case of the flying object described. For this purpose, a plurality of photosensitive part regions (pixels) are provided to detect the shift of the optical structure, output a measurement signal of the shift, and an optoelectronic shift sensor (3) including an analysis device integrated on the chip. ) In combination with the imaging optical system (1) so that the optical structure of an object at infinity is imaged onto the sensor photosensitive area with sufficient resolution to detect image shifts. It only has to be arranged in.

  As illumination, ambient light may be provided. The contrast required for shift-scan may arise from the surface of the imaged object or from the contour of one or more objects (group) with respect to the background.

  In contrast to the operation of an optical mouse, the object may be smaller than the optical capture range of the device, which range is determined by the size of the photosensitive sensor area, the focal length and the scanning distance. The object may be illuminated by light necessary for scanning.

  Here is a list of possible applications as the scanned object moves. Applications are those that detect the presence of moving objects, such as automatically opening and closing gates for incoming vehicles or visitors, as opposed to known optical sensors such as photoelectric relays or light baffles. If the background is not structured, the search range is not limited. If the background is unstructured, the detected object may be significantly smaller than the optical detection range, which is determined by the size of the photosensitive sensor area.

  Further variations include non-contact measurement of the speed or rotation of a moving object or measurement of the distance to the moving object. For example, the sensor may be directed to a transfer material placed on a conveyor belt. If the transfer speed is known from the optically measured movement speed, a measurement of the height of the transfer material is obtained. That is, the reciprocal thereof is calculated.

  In addition, continuous materials can be analyzed for error degrees of freedom. Depending on the surface condition and illumination, the error can be detected as a detected structure or as a missing structure. In contrast to commonly used image processing systems, costs can be significantly reduced.

  In addition, the velocity of the liquid or gas can be measured by suspended particles carried with it.

  As a further application, a visual representation can be obtained in response to the movement of the observer, for example a display window. For this purpose, the observer is optically scanned by the device. By scanning the observer's face, the known tracking (head tracking) problem associated with 3D image representation can be solved in a cost effective manner. The measurement signal of the shift sensor can be used as a measure of the shift in head and / or actual eye position. This signal controls the perspective of visual recognition of the image seen by the observer to match the movement of the observer's head, as well as stereo optics, so as to be incident on the correctly positioned eye even when the head moves. It is both possible to control the exit direction of exit pupils provided in separate eyes in the case of a visual image representation.

  In addition to the sensor shift signal, further measurement signals obtained from the sensor, such as illumination values or contrast quality values, can be analyzed as well. This is advantageous for detection of transferred material and machine control.

  Examples of applications in the case of moving sensor devices include robots and vehicles, and also include automatic steering, especially automatic steering with an adjustment control loop, distance detection and collision avoidance. Again, movement is measured selectively and quantitatively, or only the presence of an object such as an obstacle is detected. In the case of a lateral line of sight, the lateral distance, for example with respect to the wall, can be measured and adjusted. Thereby, as an actuator, steering can be controlled and it can be steered along an obstacle.

2 shows an example of a first embodiment for stabilizing a helicopter. Fig. 4 shows an example of a second embodiment for measuring and stabilizing rotation about a vertical axis. Fig. 5 shows an example of a third embodiment for measuring vertical velocity or distance to an object. The example of 4th Embodiment which performs the combination measurement of several degrees of freedom is shown. The block diagram of the adjustment control of the example of 1st embodiment is shown.

Claims (13)

An optical sensing system that measures the motion and / or position of a machine-controllable or drive-equipped vehicle in at least one of six spatial degrees of freedom,
The system
Optical imaging means mounted on the vehicle and suitable for projecting an image section of the outside world onto an image plane;
An optoelectronic shift sensor provided on a common substrate, and a plurality of photosensitive partial areas (pixels) and an electronic evaluation circuit that detects a shift of the pixel image and outputs a measurement signal of the shift;
The optical sensing system, wherein the optical imaging means is adapted and configured to project an optical structure of an object at infinity onto the shift sensor. The system of claim 1, wherein as the shift sensor, wherein the sensor chip used in at least an optical mouse is used.   The system according to claim 1, wherein the shift sensor has an output function of outputting a shift measurement value in the form of an incremental shift value or a shift jump.   The system of claim 1, wherein the shift sensor has less than 1,500 pixels, and pixel readout and evaluation speeds exceed 120 Hz.   In the electronic evaluation process, from the output signal of at least one measurement coordinate obtained from the shift sensor, the following measurement value, shift amount by the sum of incremental shifts, measured shift speed, rate of change of speed, measured movement The system of claim 1, further comprising an electronic device optimally adapted to obtain at least one measurement of the occurrence of.   An electronic device that is optimally adapted to convert a measurement signal output from the shift sensor into a measured value of the speed of the shift, the electronic device taking into account the direction of the shift as a leading indicator 6. The system according to claim 5, wherein a frequency of shift increment or shift jump output from the shift sensor is measured.   An electronic (control) device that generates a control value from an actual measurement value derived from the measurement signal of the shift sensor, the electronic (control) device controls the flying object in order to stabilize the flight of the flying object The system according to claim 1, wherein the system is connected to an actuating element. At least one gyroscope,
The system of claim 1, comprising: an electronic circuit adapted to compensate and mix signals derived from the gyroscope and the shift sensor. The system according to claim 1, characterized in that the flying object, provided in helicopter or other rotary wing aircraft. A photo-electronically generated measurement value output by the shift sensor is at least partially used as an actual measurement value, further comprising a feedback loop for generating a control value from the measurement value;
The system according to claim 7, wherein the control value is defined in proportion to and in inverse proportion to a speed deviation, and controls at least a roll motion.   The system of claim 10, wherein the feedback loop controls a roll tilt angle.   At least two shift sensors are used, their main line of sight resulting from optical imaging is spread by a predetermined angle, and their sensor signals are analyzed in relation to each other The system of claim 1. Remotely controllable flying vehicle, characterized by a system according to any one of claims 1 to 12.

Images